The year we woke up to AI and the need for strong space launch governance
By Ben Iannotta|December 2023
The world’s largest rocket taking off was an irresistible choice for the cover of this issue, but not just because of its visual intrigue. Much of NASA’s multibillion-dollar Artemis lunar plan hinges on Starship’s success, and Elon Musk can’t colonize Mars without it. But what’s most significant is Starship’s forcing function in an age of unprecedented commercial innovation and experimentation by SpaceX and others. Policymakers, regulators and SpaceX now have no choice but to find the right balance between, on one hand, public safety and environmental stewardship and, on the other, technical progress. The progress in the November launch was achieving stage separation. SpaceX celebrated this as a positive step in its “rapid, iterative development approach,” but FAA ordered a mishap investigation due to the loss of the two stages. A question is whether the government will be duty-bound to limit how many “mishaps” SpaceX is permitted over public waters.
As impactful as the Starship launches were, that’s not the only development for which 2023 will be remembered. Over a span of months, we witnessed artificial intelligence and machine learning “enter our collective consciousness,” as NASA’s William C. Johnson puts in the Human-Machine Teaming article. If you peruse Johnson’s piece and others, you’ll see how aerospace technologists are beginning to incorporate AI and ML into their work. That’s not to say that AIAA members are jumping into this revolution without pointing out the limits of the technology as it exists today. See, for example, the letter to the editor reacting to Moriba Jah’s column “Homo machina” and our “The elusive fully autonomous airliner” feature by Jon Kelvey, from the October issue.
Even Starship has a connection to AI in the following sense: Its creator, Elon Musk, harbors fears deeper than many of us knew over whether the technology will be put to good or ill. Walter Isaacson’s biography, “Elon Musk,” released in September, recounts how at Musk’s birthday party in 2013, Musk and Larry Page of Google fame sparred over whether it would matter if machines rendered people as irrelevant or even made us extinct. “Well, yes, I am pro- human,” Musk said. “I [expletive] like humanity, dude.”
For Musk and I’m sure many AIAA members, the topics of automation, robotics, AI and ML are no longer just futuristic concerns. They’re learning to apply the tech. Isaacson recounts a rough lesson Musk learned about the limits of automation, at least as it existed in 2017. One night at Tesla’s Gigafactory for batteries, Musk and his “posse” discovered that humans were still better at some manufacturing tasks, prompting Musk to “de-automate” some processes to meet his heady productivity goal.
Whether it’s Starship or production in a Tesla factory, this year reminds us that, for at least awhile longer, the only way to perfect a technology will be to fly it — figuratively or literally. Consider the XQ-58A Valkyrie, an uncrewed military demonstrator. Though a Valkyrie was first flown in 2019, this year one did so under the control of AI algorithms for the first time. Perhaps fighter pilots will be among the first of us to be supplanted by machines.